Biomedical Researchers Should be Taught Statistics Differently than Biostatisticians in Training: Illustration of a Module within a Clinical Research Seminar Course

Abstract

The predominant model for biomedical research is team science. Two critical members of the team are the clinical investigator and the biostatistician. Typically, the biostatistician performs statistical analyses and the clinical investigator interprets the results. Clinical investigators have different background and interests than biostatisticians, and should be taught statistics differently. Concepts should be phrased in plain language, illustrations should replace mathematical derivations, and underlying statistical concepts should be explicitly named. Consistent with basic principles of constructivism, clinical investigators and biostatisticians will (and should) have different but overlapping mental maps of statistics. Our approach is illustrated through the description of a module within a research seminar course for clinical investigators.